依據梵文佛教術語的音譯論中古漢語母音的長短問題

Vowel Length in Middle Chinese Based on Buddhist Sanskrit Transliterations

本文的主要目的在推翻Pulleyblank（1984）、Kalgren（1922）、及Chao（1940）認為中古漢語有長母音之論點。主要的反證有下列三點：首先，在Chen（2000）對漢語中梵文外來語之研究，以及羅常培（1963） 所列出四十九根本字諸經譯文異同表中，譯經者在音譯梵文字母及外來語時，並沒有使用不同之母音，來對譯梵文長母音和短母音。其二，筆者從Soothill 和Hodou’s（1968）所寫的《漢英佛學大辭典》（A Dictionary of Chinese Buddhist Terms）收集了約三千個梵文佛教術語，來研究其對譯的漢語語料。因每一個梵文術語都曾被多次翻譯，故共有七、八千筆漢語對音資料，其中共有1426個漢字，用來對譯梵文長母音音節。研究結果顯示，所有常用來對譯梵文長母音的漢字，也都用來對譯梵文短母音。其三，從資料中也發現即使在同一梵文術語中，有相同性質的長母音和短母音，翻譯者也使用相同的漢字來對譯。基於以上三點，筆者認為中古漢語應該是沒有長短母音之對立，不然佛經的翻譯者在對音時，應會使用長母音來對譯梵文長母音，短母音來對譯梵文短母音，不至於毫無分別地使用相同的漢字。

The goal of this paper is to disprove Pulleyblank (1984), Kalgren (1922) and Chao (1940)’s assumption that there were long vowels in Middle Chinese. The main arguments are: first, in Chen’s (2000) study of Chinese loanwords from Sanskrit, and in Lo’s (1963) list of the forty-nine Sanskrit sounds transliterated from sixteen Buddhist works, it was found that there was no distinction between short vs. long vowels. Moreover, in data collected from Soothill and Hodous’ A Dictionary of Chinese Buddhist Terms (1968), it was found that all the characters frequently used to transcribe long vowels were also used to transcribe short vowels. Had there been a short-vs.-long-vowel distinction in Middle Chinese, different characters should have been used. Finally, I have found that even within the same words there were short and long vowels which were transliterated with the same Chinese characters.